Mapped to a shared offline runtime family and usable with a recording or compatible ESP32 CSI stream.
Intended capability
Pilot this intended outcome through the shared Zone dwell family, then validate it against site-specific ground truth: Watches each charge-dock lane for docking that stops working: approach-retreat cycles that never end in stillness at dock range — dock-seek fugue — plus climbing approach-per-dock ratios and lengthening final-approach settle times, the drift that precedes the fugue. The cone verifies the bay physically empty, so retries against a verified-empty dock are fugue on physics alone; declared dispatch and charge state only refine the judgment. Fugue's cost is nonlinear — enough failed approaches strand a robot mid-aisle and convert quiet waste into blockage and manual recovery — and this row exists to see it building before the stranding.
This describes the intended outcome. Readiness is shared-family mapped, evidence is class B, and a catalog mapping or recording is not proof of this outcome at a real site.
Solution blueprint
See the environment before installing it.
This exact kit is one of 191 first-class designs. It includes geometry, objects, nodes, wording, scenarios, installation, limitations, and catalog-bound readiness.
No rendered revision is available yet.
Bundled recording
A related Fleet scene
This is one recorded vertical scenario. It is not separate validation of every capability in the catalog.
01 The physics
An LD2410C aimed down the dock lane fits the fixture exactly: one lane, one target at a time, range-gated to the bay, so the sensor's no-azimuth, no-count limits cost nothing here. The sensor's human-tuned target logic is bypassed: tracking runs on engineering-mode per-gate energies with our own thresholds, because the firmware's presence logic may filter a chassis and a dead-still robot is clutter, not a designed static target. Approach, retreat, and terminal stillness at dock range are read from per-gate energy transitions; empty-vs-occupied bay is a clutter-profile difference — a parked chassis shifts the static gate baseline — benched, not assumed. CSI counts charge-zone entries as corroboration; audio band energy at the dock notes presence or absence of engagement transients as a supporting channel, never the verdict.
02 Shared processing path
Engineering-mode per-gate energy stream -> range-track segmentation per approach on own thresholds -> terminal-stillness-at-dock-range test -> per-lane approach-per-dock ratio + settle-time EWMA -> bay-occupancy verification by clutter-profile differencing on the static gate baseline -> fugue candidate on repeated approach-retreat against a verified-empty bay -> reasoning-tier twin check against declared dispatch interrupts (read-only) -> dock-health trend + fugue findings.
This is a capability design path. Components may be shared with other catalog entries; it is not presented as a unique algorithm.
03 Validation plan
Bench and site protocol: gate-energy response characterized across real dock approach speeds on an AMR test rig; clutter-profile empty-vs-occupied differencing benched against ground truth; scripted docking cycles with induced failures — occluded fiducial, offset dock — scoring approach segmentation and terminal-stillness detection; then a 30-day per-lane site soak for drift-trend stability.
04 Commercial hypothesis
'Dock Watch' — for the fleet maintenance owner who keeps finding robots stranded at 04:00: the per-lane record of docking health, and the fugue alarm that fires early in the failure sequence — before the stranding, not after it.